Experimental neutron spectroscopy data visualization: Adaptive tessellation algorithm

Abstract

We report on an adaptive binning approach designed for data visualization within scientific disciplines where counting statistics are expected to follow Poisson distributions. We envisage a wide range of applications stemming from astrophysics to the condensed matter sciences. Our main focus of interest concerns, however, neutron spectroscopy data from single-crystal samples where signals span a four-dimensional space defined by three spatial coordinates plus time. This makes widely used equal-width binning schemes inadequate since physically relevant information is often concentrated within rather small regions of such a space. Our aim is thus to generate optimally binned data sets from one-dimensional to three-dimensional volumes to provide the experimentalist with enhanced ability to carry out searches within a four-dimensional space. Several binning algorithms are then scrutinized against experimental as well as simulated data.